Sarcouncil Journal of Economics and Business Management

Abstract: Audit quality is a vital basis of transparency and investor trust within the U.S. capital markets, which are regulated by the Public Company Accounting Oversight Board (PCAOB) and the Securities and Exchange Commission (SEC). The traditional methods of auditing that were bound by manual sampling and human influence cannot safely cope with the up-to-date data-intensive corporate setup. This review analyzes the application of machine learning (ML) devices, deep learning (DL), and natural language processing (NLP) in improving audit supervision of publicly traded companies. The research aim is to determine the effectiveness of AI in terms of compliance with regulatory requirements, enhanced fraud detection, and increased investor confidence. This conceptual paper reviews recent literature and analyzes two case studies. The first case study examines the successful implementation of PwC’s GL.ai, which has enabled real-time anomaly detection and enhanced compliance reporting. The second is the unsuccessful early implementation of Clara AI at KPMG, which was marred by data bias and obscurity of the algorithm. Results indicate that AI implementation improves the accuracy of audits, speeds up the process of compliance control, and makes risk assessment more reliable, but there are still ethical and technical issues. The research concludes that successful AI implementation with the help of explainable AI models, governance structures, and professional education can turn hurdle-retrospective verification of auditing into predictive assurance. Finally, responsibly implementing AI technologies improves the quality of audit and builds trust in the integrity and transparency of financial reporting through the U.S. capital markets.